作者: Monica Trifas , Nikolaos Bourbakis , Ming Yang , Christopher Cushing
DOI:
关键词:
摘要: With the proliferation of digital multimedia, information hiding techniques have become more attractive to researchers. High bitrate is different from watermarking: former one tries hide relatively large amount auxiliary instead just or a few verification bits. This paper presents novel Discrete Wavelet Transform (DWT) domain high algorithm. In proposed algorithm, coefficients within approximation subband one-level wavelet decomposition are grouped into vectors embed Low-frequency been chosen for due their amplitudes and corresponding smaller step size in JPEG2000 quantization. A mathematical model predict Bit-Error-Rate (BER) algorithm under compression. The has reached an initial channel capacity 1/64 bits per pixel (bpp). performance improvements, increased 1/32bpp 1/16bpp, at price slightly higher BER. Preliminary experiments indicate that robust most hidden survives